•  
  •  
 

UC Law SF International Law Review

Abstract

This article examines how the integration of artificial intelligence-enabled decision-support systems (AI-DSS) into military conflict complicates the attribution of individual criminal responsibility for war crimes under the Rome Statute. Unlike lethal autonomous weapons systems which are intended to supplant human agency, AI-DSS augments human decision-making across the targeting cycle while simultaneously introducing opacity and bias. These characteristics risk facilitating indiscriminate and disproportionate attacks, undermining t4he principles of distinction and proportionality under international humanitarian law. Through a case study of Israel’s deployment of systems such as “the Gospel,” “Lavender,” and “Where’s Daddy?” in Gaza, this article illustrates how algorithmic mediation obscures the attribution of actus reus and mens rea, creating a potential “responsibility gap” through which commanders may evade liability for war crimes.

To address this critical lacuna, the article advances a doctrinal innovation: a “rebuttable presumption of recklessness” for commanders who authorize or fail to prevent the use of high-risk AI-DSS. Under this framework, liability attaches unless the accused demonstrates due diligence in deploying systems anticipated to generate significant risks of unlawful harm. This proposal balances fidelity to fault-based liability with the practical need to overcome evidentiary barriers instantiated by algorithmic opacity and emergent bias. By shifting the evidentiary burden to those with superior responsibility, the presumption ensures that accountability is preserved as warfare becomes increasingly mediated by AI. Ultimately, this article argues that such an evolution in international criminal law is necessary to pierce the algorithmic veil of modern conflict.

Download

Share

COinS